Abstract
Objective:
Recent data indicate social determinants of health (SDOH) have a great impact on prevention and treatment outcomes across a broad variety of disease states, especially cardiovascular diseases. The Area Deprivation Index (ADI) is a validated measure of neighborhood level disadvantage capturing key social determinate factors. Abdominal aortic aneurysm rupture (rAAA) is highly morbid, but also preventable through evidence-based screening. However, the association between rAAA and SDOH is poorly characterized. Our objective is to study the association of SDOH with rAAA and screening age.
Methods:
This was a retrospective study of patients who underwent operative repair of a rAAA at a multi-hospital healthcare system (2003-2019). Deprivation was measured in ADI (scale 1-100), grouped into quintiles for simplicity, with higher quintiles indicating greater deprivation. Patients with the highest quintile ADI (89-100) were categorized as the most deprived. We investigated the association between neighborhood deprivation with the odds of i) undergoing repair for rAAA before screening age 65, and ii) undergoing endovascular aortic repair (EVAR) using logistic regression, sequentially modeling non-modifiable then both non-modifiable and modifiable confounding variables.
Results:
632 patients met inclusion criteria (age 74.2 ± 9.4 years; 174 [27.6%] women; 564 [89.2%] White, ADI 66.8 ± 22.3). Those from the most deprived neighborhoods (n=118) were younger (71.7 ± 10 vs 74.8 ± 9.2 years; p=0.002), more likely to be female (36% vs 26%; p=0.031), more likely to be Black (5.9% vs 0.4%; p=0.007), and fewer underwent EVAR (28% vs 39.5%; p=0.020) compared with those from other neighborhoods. On sequential modeling, residing in the most deprived neighborhoods was associated with undergoing rAAA repair before age 65 after adjusting for non-modifiable factors (OR 2.02; 95% CI 1.39-2.95; p<0.001), and non-modifiable as well as modifiable factors (OR 2.22; 95% CI 1.56-3.16; p<0.001). Those in the most deprived neighborhoods had a lower odds of undergoing EVAR compared to open repair after adjusting for non-modifiable factors (OR 0.64; 95% CI 0.41-0.98; p=0.042), and non-modifiable as well as modifiable factors (OR 0.61; 95% CI 0.37-0.99; p=0.047).
Conclusions:
Among patients who underwent rAAA repair, residing in the most deprived neighborhoods was associated with greater adjusted odds of presenting under age 65 and receiving an open repair. These neighborhoods represent tangible geographic targets that may benefit from a younger screening age, enhanced education, and access to care. These findings stress the importance of developing strategies for early prevention and diagnosis of cardiovascular diseases among patients with disadvantageous SDOH.
Keywords: Neighborhood deprivation, ruptured abdominal aortic aneurysm, disparities, health equity, social determinants of health
TABLE OF CONTENTS SUMMARY
In this retrospective study of 632 patients who underwent rAAA repair, residing in the most deprived neighborhoods was associated with presenting under the screening age of 65 years compared to those from less-deprived neighborhoods. These regions may represent areas for targeted interventions such as enhanced education, screening, and follow-up.
INTRODUCTION
Social determinants of health (SDOH) such as economic stability, education, housing quality, healthcare, and community shape population health and impact health outcomes. The Centers for Medicare and Medicaid Services recently indicated that SDOH are more than an adjunct to the health care ecosystem, but are driving the definition of healthcare itself.1 Many individual components of SDOH and their impact on health outcomes have been studied in patients with vascular disease. For example, sociodemographic characteristics like income, educational attainment, and race are independently associated with cardiovascular outcomes. People with low socioeconomic status and racial minority groups present with more severe cardiovascular diseases which manifest at younger ages compared to other, majority groups.2,3 Among patients with abdominal aortic aneurysms (AAA), low household income and education level as well as Black race have been described as independent predictors of presenting with a ruptured abdominal aortic aneurysm (rAAA).4,5 Similarly, Black race, low socioeconomic position, and Hispanic ethnicity have been shown as associating factors with a lower odds of receiving an endovascular aortic repair (EVAR) over open repair; however, this remains controversial.6,7
There remain some discrepancies between societal and United States Preventative Services Task Force (USPSTF) issued guidelines for AAA screening and elective repair in the prevention of AAA-related mortality. Both the USPSTF and Society of Vascular Surgery (SVS) recommend one-time AAA screening among those 65 to 75 years of age in the setting of known tobacco use. The SVS maintains this guideline for both men and women (Grade 1A)8, while the USPSTF limits their recommendation to men (Grade B) and recommend against screening such women without a concomitant AAA family history (Grade D).9 However, both sets of societal guidelines were informed by European studies, with mostly male participants, with less than 20% reporting sociodemographic characteristics such as race or other SDOH. Despite consistent guidelines targeting the onset age for screening, a recent study found that 20% of patients with rAAA undergoing emergent repair were under the age of 65 and therefore never had an indication for screening.10 Others have suggested that further targeted screening in those at highest risk may be particularly high-yield to prevent disparities seen later in the disease course.7
Area deprivation index (ADI) is a validated measure of socioeconomic disadvantage at the neighborhood level, and has demonstrated utility in the assessment of disparate outcomes across a broad variety of disease states.11-13 The national ADI is based on home address and is derived from 17 variables that are related to education, employment, housing quality, and poverty.14 This tool estimates SDOH and allows for the investigation into geographic regions that may be good populations of interest for interventions such as enhanced screening campaigns or a younger screening age. In this cohort study, we sought to investigate the relationship between neighborhood deprivation using the ADI, and the odds of presenting with a rAAA prior to the current screening age of 65 and secondarily the use of EVAR when compared to open repair. We hypothesized that ADI would be independently associated with presenting under age 65 and undergoing open AAA repair.
METHODS
This was a single healthcare network, multi-hospital, retrospective cohort study of patients who underwent repair for a rAAA from 2003 through 2019. As a quaternary referral center, patients with a rAAA are transferred both within and from outside our healthcare network to our hospitals from Southwestern Pennsylvania and the surrounding states. Patients were identified through retrospective review of institutional operative records that contain detailed accounts of all operations performed by the Division of Vascular Surgery. A retrospective review of the electronic health record (EHR) confirmed the diagnosis of rAAA and receipt of repair. Patients were excluded if they i) did not have a listed home address allowing ADI assignment, ii) experienced a traumatic or iatrogenic rupture, or iii) had a known vascular genetic disease (i.e., Marfan Syndrome). Patients who had undergone surgery for rAAA more than once within our network were only included in the first encounter. This study was in compliance with principles outlined in the Declaration of Helsinki and approved by the University of Pittsburgh Human Research Protection Office (STUDY19070316). All collection, analyses (STATA 17.0; Stata corp.), and presentation of data are in compliance with the Strengthening The Reporting of Observational Studies in Epidemiology (STROBE).15
Patient characteristics
Our data includes patient baseline demographics (i.e., age, sex, self-reported race and ethnicity), comorbidities (i.e., hypertension, chronic obstructive pulmonary disease [COPD], coronary artery disease, smoking status), as well as procedural information (i.e., EVAR vs open repair). Smoking status was classified as ever-smoker (i.e., current or prior tobacco use) vs. non-smoker. History of known AAA was determined from detailed chart review and was considered present if patient had stated knowledge of their AAA on presentation, or imaging or documentation of AAA in the EHR prior to rupture encounter. Missingness of data collected for the entire cohort was quantified (Table sI).
Among patients too young for recommended screening (<65 years of age), AAA diagnosis was classified as either identified through screening, incidentally noted on prior imaging, or diagnosed on imaging performed for evaluation of symptoms (e.g., abdominal or back pain, pulsatile abdominal mass). Time in days from diagnosis to operative intervention was calculated for patients who had a diagnosis for at least one day prior to repair. The presence of a listed primary care physician (PCP) in the EHR was noted, as well as the utilization of primary care services within the two years prior to rupture. Preventative healthcare utilization was considered receipt of care from any non-surgical physician or provider that was actively managing chronic medical conditions (example: internists, family physicians, cardiologists). Insurance was classified as either present or absent at the time of repair. Active preoperative prescriptions for select medications (including statin, aspirin, and metformin), was determined based on listed outpatient medications at the time of repair or relevant history in the admission records.
Neighborhood Deprivation
The home address for each patient was collected from the EHR between October 2020 through February 2021. Each home address was simultaneously linked to a corresponding state (range 1-10) and national (range 1-100) ADI through the Neighborhood Atlas website,14 reflecting neighborhood level deprivation. A higher ADI correspond with a higher level of deprivation. Neighborhood is defined as a census block group, a geographically compact area which typically contains between 600 to 3000 individuals. Each census block group is assigned a number that reflects the level of deprivation and is normally distributed compared to other neighborhoods in the country (for national ADI), or the individual state (for state ADI). The distribution of ADI in our cohort clustered to the right, and therefore for simplicity and clarity of concept we split our cohort into five groups of relatively equal sizes to reflect the different quintiles of deprivation (ADI is represented as a whole number, so some groups have more than others), as has been done in previous studies utilizing the ADI.11-13 The highest quintile was then categorized as the most deprived, while the remaining four quintiles were categorized as less-deprived.
National ADI was used in our study due to our large catchment area, although the state ADI is also reported. National ADI is a standardized estimation of neighborhood deprivation based on census data as well as data from the American Community Survey (ACS). The ACS contains 5-year estimates, so the version we used in this study (2019 ADI) contains average values from 2015-2019.Variables included in the measure reflect the theoretical domains of income, education, employment, and housing quality. Geographic representation of this data was created by color coding national ADI data onto a map of census block groups (IPUMS National Historical Geographic Information System: version 16.0, 2021) using open source geospacial software (QGIS Geographic Information System, 2021).16,17
Outcome of interest
Our primary outcome of interest was presenting with a rAAA before the age of 65 years. We chose this as our primary outcome to investigate whether deprived neighborhoods may represent geographic targets for enhanced screening, or educational campaigns in relation to aneurysmal disease and rupture prevention. Secondary outcomes included receipt of an EVAR compared to open repair.
Statistical analysis
Patient characteristics including demographic, comorbid conditions, and operative details were assessed for each quintile of ADI overall and among those under the age of 65 years. We presented normal continuous variables as mean (±standard deviation), skewed continuous variables as median (interquartile range), and categorical variables as frequency (percentage). To evaluate baseline differences between cohorts, we used Student t test for normal continuous variables, Kruskal-Wallis test for skewed continuous variables, and chi-square test for categorical variables. To explore the association between highest level of deprivation and our outcomes of interest, we sequentially modeled our analysis adjusting first for non-modifiable characteristics (i.e., age, sex, race, year of repair), followed by adjusting for non-modifiable as well as modifiable characteristics (i.e., tobacco use, hypertension, chronic obstructive pulmonary disease [COPD], coronary artery disease). This was done to investigate how modifiable and non-modifiable confounding variables associate with our outcomes, as modifiable characteristics and underlying chronic medical conditions may be related to SDOH.18,19 All analyses were done using a cluster-robust variance estimator to account for correlation at the level of the operating hospital (n=8 hospitals). Our primary hypothesis (i.e., age of presentation is associated with ADI) and secondary hypotheses (i.e., repair type is associated with ADI) as well as select variables used for sequential modeling (i.e., age, sex, race, tobacco use, diabetes, and year of repair) were determined a priori. A p-value<0.05 determined a statistically significant difference between groups with no correction for multiple comparisons.20 Our included sample size has 90% power to detect a 14% difference between groups. This is greater than the clinically meaningful differences estimated to be 10% in the primary outcome, with the less deprived groups estimated to have 20% with the primary outcome, and the most deprived group estimated to be at 30%.2,7
Sensitivity analysis
To interrogate our model and confirm the robustness of our results, we performed three distinct sensitivity analyses. First, we evaluated our primary and secondary outcomes using a definition of most deprived based on quartiles of ADI, instead of quintiles. Second, we used a definition of most deprived based on sextiles (n=6) of ADI. Third, among patients residing in Pennsylvania, we completed our analysis using the state ADI. Fourth, to evaluate for a dose-response association, we used ordinal regression to evaluate our primary outcome with ADI quintiles. In each of these separate analyses, we performed the same modeling as our primary analysis adjusting for non-modifiable factors then both for non-modifiable and modifiable factors.
Results
We identified 632 patients who underwent rAAA repair met study criteria (age 74.2 ± 9.4 years; 174 [27.6%] women; 564 [89.2%] White, ADI 66.8 ± 22.3, Figure 1). Overall, 120 (19.0%) were under the age of 65 and 236 (37.3%) underwent EVAR. The distribution of ADI for this cohort clustered to the right, toward more deprived areas (Figure 2). Those in the most deprived neighborhoods (n=118; ADI 89-100) were younger (71.7 ± 10 vs 74.8 ± 9.2 years; p=0.002), with nearly a third undergoing rAAA repair before the screening age (28.0% vs 16.9%; p=0.006), compared to those who resided in the less-deprived neighborhoods (n=514; ADI 5-88) (Figure 3). They were also more likely to be female (36% versus 25%; p=0.031), and more likely to self-identify as Black (5.9% vs 0.4%; p=0.007). There were no significant differences in comorbid conditions, including history of tobacco use, between groups. And finally, fewer patients who resided in the most deprived neighborhoods underwent EVAR (28% vs 39.5%; p=0.020) compared to those in the less-deprived neighborhoods (Table I; Figure 4).
Figure 1. – Consort diagram of study cohort.
Abbreviations: rAAA, ruptured abdominal aortic aneurysm
Figure 2. – Distribution of Area Deprivation Index by quintile.
Abbreviations: ADI, area deprivation index; rAAA, ruptured abdominal aortic aneurysm
Figure 3. – Age by Area Deprivation Index.
Age and ADI are inversely related with a higher ADI (more deprived) corresponding to a younger age on average. Dots represent individuals who underwent repair for a ruptured abdominal aortic aneurysm, line represents the best fit line, vertical gray lines represent the maximum ADI in each quintile.
Table I. –
Characteristics of patients with ruptured AAA stratified by neighborhood disadvantage quintiles
| Variables | Less-Deprived | Most Deprived | Pa | |||
|---|---|---|---|---|---|---|
| Quintile 1 (n=129) |
Quintile 2 (n=132) |
Quintile 3 (n=123) |
Quintile 4 (n=130) |
Quintile 5 (n=118) |
||
| Demographics | ||||||
| ADI (national) | 33.1 ± 10.3 | 55.8 ± 5.2 | 70.0 ± 3.1 | 82.5 ± 3.6 | 95.2 ± 3.6 | <0.001 |
| ADI (state)b | 2.5 ± 1.1 | 5.2 ± 1.0 | 6.7 ± 0.7 | 8.3 ± 0.8 | 9.7 ± 0.6 | <0.001 |
| Age (years) | 75.8 ± 9.4 | 76.2 ± 8.8 | 73.7 ± 8.7 | 73.3 ± 9.7 | 71.7 ± 10.0 | <0.001 |
| Under age 65c | 20 (15.5%) | 16 (12.1%) | 21 (17.1%) | 30 (23.1%) | 33 (28.0%) | 0.012 |
| Female sex | 30 (23.3%) | 40 (30.5%) | 30 (24.4%) | 32 (24.6%) | 42 (35.6%) | 0.15 |
| Race | ||||||
| White | 118 (91.5%) | 119 (90.2%) | 112 (91.1%) | 119 (91.5%) | 96 (81.4%) | 0.001 |
| Black | 0 ( 0.0%) | 0 ( 0.0%) | 0 ( 0.0%) | 2 ( 1.5%) | 7 ( 5.9%) | |
| Asian | 2 ( 1.6%) | 0 ( 0.0%) | 0 ( 0.0%) | 1 ( 0.8%) | 0 ( 0.0%) | |
| Other / Unknown | 9 ( 7.0%) | 13 ( 9.8%) | 11 ( 8.9%) | 8 ( 6.2%) | 15 (12.7%) | |
| Comorbid conditions d | ||||||
| Tobacco usee | 85 (73.9%) | 84 (74.3%) | 84 (84.8%) | 92 (82.9%) | 84 (84.0%) | 0.097 |
| Hypertension | 88 (73.3%) | 92 (73.0%) | 73 (67.0%) | 81 (68.6%) | 77 (70.6%) | 0.79 |
| Hyperlipidemia | 42 (35.3%) | 36 (28.8%) | 43 (39.4%) | 42 (35.9%) | 32 (29.9%) | 0.40 |
| Coronary artery disease | 51 (42.9%) | 46 (36.8%) | 43 (39.4%) | 41 (35.0%) | 44 (41.5%) | 0.73 |
| Congestive heart failure | 11 ( 9.3%) | 9 ( 7.3%) | 9 ( 8.3%) | 8 ( 7.0%) | 12 (11.2%) | 0.79 |
| COPD | 35 (29.4%) | 29 (23.2%) | 27 (24.5%) | 31 (26.3%) | 35 (32.7%) | 0.49 |
| Diabetes | 10 ( 8.5%) | 25 (20.0%) | 23 (21.3%) | 21 (18.1%) | 22 (20.6%) | 0.062 |
| Chronic kidney disease | 10 ( 8.4%) | 12 ( 9.6%) | 8 ( 7.4%) | 9 ( 7.7%) | 12 (11.2%) | 0.86 |
| Cancer | 24 (20.9%) | 19 (15.1%) | 14 (13.0%) | 12 (10.4%) | 15 (13.9%) | 0.24 |
| Prior known history of AAA | 42 (34.4%) | 46 (36.8%) | 24 (21.4%) | 46 (38.0%) | 45 (40.9%) | 0.022 |
| Operative details | ||||||
| EVAR | 59 (45.7%) | 43 (32.6%) | 54 (43.9%) | 47 (36.2%) | 33 (28.0%) | 0.018 |
| Year | ||||||
| 2003 – 2007 | 27 ( 21%) | 34 ( 25.8%) | 28 ( 22.8%) | 35 ( 26.9%) | 28 ( 23.6%) | 0.014 |
| 2008 – 2012 | 44 ( 34.2%) | 31 ( 23.4%) | 38 ( 30.9%) | 32 ( 24.7%) | 37 ( 31.3%) | |
| 2013 – 2017 | 37 ( 28.8%) | 44 ( 33.2%) | 41 ( 33.3%) | 47 ( 36.2%) | 35 ( 29.7%) | |
| 2018 – 2020f | 21 ( 16.3%) | 23 ( 17.4%) | 16 ( 13%) | 16 ( 12.3%) | 18 ( 15.2%) | |
Abbreviations: AAA, abdominal aortic aneurysm; ADI, area deprivation index; COPD, chronic obstructive pulmonary disease; EVAR, endovascular aortic repair
Categorical variables are number (%) and continuous variables are mean ± SD or median (IQR)
Chi-square and ANOVA were used to test differences among the 5 quintiles and considered significant with a P-value <0.05.
Pennsylvania specific area deprivation index, a measure of deprivation based on neighborhoods within the state of Pennsylvania alone (N=583), normalized based on census block group, and measured on a scale from 1 (least deprived) to 10 (most deprived).
Current screening age according to SVS practice guidelines is 65 for men and women with a history of smoking.6
At the time of repair
Current or prior tobacco use
Through June, 2020
Figure 4. – Neighborhood Deprivation Heat Map with Patient Home Address Overlay.
Abbreviations: ADI, area deprivation index; EVAR, endovascular aortic repair. This map depicts the home address (points on map) for patients who have undergone open (circle), and endovascular (triangle) repair and for patients who are under (pink) and over (green) age 65. Blue indicates a lower ADI representing less deprivation, whereas red represents a higher ADI representing more deprivation. Observe that most patients under 65 reside in areas that appear red on the heat map.
Outcomes stratified by Area Deprivation Index
Residing in the most deprived neighborhoods was associated with undergoing repair for rAAA before the screening age after adjusting for non-modifiable patient characteristics (OR 2.02; 95% CI 1.39-2.95; p<0.001) and adjusting for non-modifiable as well as modifiable variables (OR 2.22; 95% CI 1.56-3.16; p<0.001) with more deprived patients having a higher odds of presenting prior to age 65 when compared with the less-deprived groups (Table II).
Table II. –
Sequential logistic regression modeling association between neighborhood deprivation and i) presenting under the screening age of 65 and ii) endovascular aortic repair
| Age under 65 yearsa, b | Endovascular aortic repairc | |||||||
|---|---|---|---|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | OR (95% CI) | P | |
| Most deprivedd | 2.02 (1.39-2.95) | <0.001 | 2.22 (1.56-3.16) | <0.001 | 0.64 (0.41-0.98) | 0.042 | 0.61 (0.37-0.99) | 0.047 |
| Non-modifiable confounding variables | ||||||||
| Age under 65 | n/a | n/a | n/a | n/a | 0.72 (0.48-1.10) | 0.13 | 0.78 (0.50-1.22) | 0.28 |
| Female sex | 0.50 (0.35-0.70) | <0.001 | 0.59 (0.44-0.78) | <0.001 | 0.70 (0.50-0.99) | 0.044 | 0.61 (0.40-0.95) | 0.027 |
| Non-white race | 2.62 (1.00-6.89) | 0.05 | 2.80 (1.06-7.15) | 0.037 | 0.41 (0.09-1.82) | 0.24 | 0.49 (0.12-2.08) | 0.34 |
| Year | 1.02 (0.99-1.05) | 0.25 | 0.99 (0.96-1.03) | 0.65 | 1.18 (1.11-1.26) | <0.001 | 1.20 (1.11-1.30) | <0.001 |
| Modifiable confounding variables e | ||||||||
| Tobacco usef | 2.35 (1.37-4.03) | 0.002 | 0.54 (0.29-1.01)) | 0.052 | ||||
| Hypertension | 0.41 (0.31-0.54) | <0.001 | 1.25 (1.05-1.50) | 0.014 | ||||
| COPD | 0.73 (0.62-0.85) | <0.001 | 1.43 (0.65-3.13) | 0.38 | ||||
| Coronary artery disease | 1.50 (1.12-2.00) | 0.006 | 1.56 (1.01-2.40) | 0.043 | ||||
Sequential modeling for 1) age under 65 years old first adjusting for non-modifiable confounding variables (first row under “age under 65 years”) then controlling for non-modifiable and modifiable confounding variables (next row), and 2) use of EVAR with similar adjustments.
Abbreviations: COPD, chronic obstructive pulmonary disease; OR, odds ratio; CI, confidence interval
All models clustered by hospital where repair took place
Age was excluded from our models evaluating of the primary outcome of presenting under age 65.
P <0.05 was used as the significance level for testing the primary hypothesis. Subsequent analyses were primarily hypothesis generating and so significance levels were not corrected for multiple hypothesis testing.”
Highest quintile of area deprivation index compared to the remaining (less-deprived) patients (quintiles 1-4)
At the time of repair
Current or prior tobacco use
Those in the most deprived neighborhoods were also found to have a lower odds of undergoing EVAR after adjusting for non-modifiable confounding variables (OR 0.64; 95% CI 0.41-0.98; p=0.42), and non-modifiable as well as modifiable confounding variables (OR 0.61; 95% CI 0.37-0.99; p=0.047) compared to those from the less-deprived areas (Table II).
Patients under age 65
When we limited our cohort to those under the age of 65 (n=120), we found that a higher percentage of those in the most deprived neighborhoods were Black (15.2% vs 0%; p=0.003) compared to the less-deprived areas. Nearly a third (n=34 [28.3%]) of patients under 65 had prior knowledge of their AAA. While this was not different between ADI groups (33.3% vs 26.4%; p=0.48), the number of days from diagnosis to repair for rupture was greater for those from the most deprived areas (median days[IQR][1,188(644.0-2557.0) vs 297(4.5-1379.0); p=0.05]) compared to less-deprived areas (Table III). The presence of a designated PCP, preventative healthcare utilization, insurance status, and the use of statin, aspirin, or metformin use was not different between groups.
Table III.
Characteristics of patients under 65 by level of neighborhood disadvantage
| Less-deprived (n=87) | Most deprived (n=33)a | P | |
|---|---|---|---|
| Age (year) | 60.6 ± 3.8 | 59.3 ± 5.1 | 0.16 |
| Female sex | 15 (17.2%) | 6 (18.2%) | 0.90 |
| Race | 0.003 | ||
| White | 82 (94.3%) | 26 (78.8%) | |
| Black | 0 ( 0.0%) | 5 (15.2%) | |
| Asian | 1 ( 1.1%) | 0 ( 0.0%) | |
| Unknownb | 4 ( 4.6%) | 2 ( 6.1%) | |
| Ethnicity | 0.44 | ||
| Not Hispanic | 76 (87.4%) | 27 (81.8%) | |
| Unknownb | 11 (12.6%) | 6 (18.2%) | |
| Tobacco usec | 73 (90.1%) | 29 (87.9%) | 0.72 |
| Hypertension | 42 (50.0%) | 24 (75.0%) | 0.015 |
| Hyperlipidemia | 25 (29.8%) | 6 (18.8%) | 0.23 |
| Coronary artery disease | 36 (42.9%) | 17 (53.1%) | 0.32 |
| Congestive heart failure | 4 ( 5.1%) | 3 ( 9.4%) | 0.40 |
| COPD | 19 (22.6%) | 10 (31.2%) | 0.34 |
| Diabetes | 15 (18.1%) | 8 (25.0%) | 0.41 |
| Chronic kidney disease | 8 ( 9.5%) | 4 (12.5%) | 0.64 |
| Cancer | 6 ( 7.9%) | 3 ( 9.4%) | 0.80 |
| Prior aortic surgeryd | 9 (11.0%) | 5 (15.2%) | 0.54 |
| Prior known history of AAA | 23 (26.7%) | 11 (33.3%) | 0.48 |
| AAA diagnosis | 0.38 | ||
| Screening | 0 (0%) | 0 (0%) | |
| Incidentally noted on prior imaging | 23 (26.4%) | 11 (33.3%) | |
| AAA diagnosed on imaging for abdominal or back pain | 64 (73.6%) | 22 (66.7%) | |
| Days from diagnosis to rupture (if > 0) | 297.0 (4.5-1379.0) | 1188.5 (644.0-2557.0) | 0.050 |
| Designated PCP | 68 (81.9%) | 26 (78.8%) | 0.70 |
| Preventative healthcare utilization within 2 years prior to rupturec | 27 (32.9%) | 15 (46.9%) | 0.17 |
| Insured | 68 (82.9%) | 23 (74.2%) | 0.30 |
| Preoperative statin | 21 (29.2%) | 7 (23.3%) | 0.55 |
| Preoperative aspirin | 27 (37.5%) | 11 (36.7%) | 0.94 |
| Preoperative metformin | 6 ( 8.3%) | 1 ( 3.3%) | 0.36 |
| EVAR | 31 (35.6%) | 8 (24.2%) | 0.23 |
Abbreviations: AAA, abdominal aortic aneurysm; ADI, area deprivation index; COPD, chronic obstructive pulmonary disease; EVAR, endovascular aortic repair; PCP, primary care provider
Categorical variables are number (%) and continuous variables are mean ± SD or median (IQR)
Highest quintile of Area Deprivation Index
No mention of race or ethnicity in the medical record
Current or prior tobacco use
Including prior open or endovascular thoracic or abdominal aortic repair (n=13), or prior ascending aortic repair (n=1)
Receipt of care from any non-surgical physician or provider including internal medicine, family medicine, and select medical specialty care (for chronic medical conditions)
Sensitivity analysis
We demonstrated the robustness of our results by performing multiple sensitivity analyses. When grouping patients by first quartiles then sextiles of ADI, the association between neighborhood deprivation and the odds of presenting under age 65 was consistent in sequential modeling (Table sII). When using our different grouping methods, the association between the most-deprived neighborhoods and a lower odds of receiving an EVAR was again consistently observed in sequential modeling.
Restricting our cohort to those residing in Pennsylvania, and using exclusively the Pennsylvania ADI (n=583), the association of deprivation and presenting under the screening age was consistent on sequential modeling. Similarly, those from the more deprived neighborhoods had a lower odds of undergoing EVAR after adjusting for modifiable and non-modifiable confounding variables, and failed to meet significance when adjusting for non-modifiable confounding variables.
Evaluating our primary outcome using ordinal regression, the association of deprivation and presenting under age 65 was seen and consistent on sequential modeling (Table sIII).
Discussion
In this large retrospective cohort study of patients undergoing repair for rAAA, we observed that residing in the most deprived neighborhoods was associated with a higher adjusted odds of undergoing rAAA repair under the age of 65 and a lower adjusted odds of receiving an EVAR compared to those from the less-deprived neighborhoods.
Recommendations for screening rely on a fundamental understanding of the disease process, as well as supporting literature, financial considerations, and carefully done population-based studies.21-23 While it is well established that aneurysmal disease is related to aging, there is a paucity of data to understand the relationship between SDOH and AAA progression and ultimately rupture.24 Overall, it has been shown that approximately 20% of patients presenting with a rAAA do so prior to the current screening age of 65.10 Yet in the most deprived neighborhoods, we found that this number climbs to nearly 30%. Interestingly, we found that prior knowledge of AAA increased with quintile of ADI, and did not differ between groups for all patients under 65. Taken together, these findings imply that the most deprived neighborhoods require focused, future investigation and may be particularly fruitful populations of interest for enhanced education, access to care, and a younger screening age. Doing so would work toward preventing rupture in under resourced neighborhoods that appear to be disproportionately affected by cardiovascular disease.6
In addition to disparities in age of presentation of rAAA, we found an association between level of deprivation and the odds of receiving an EVAR, with those from the most deprived areas having a 34 to 39% lower odds of receiving an EVAR compared to those from the less-deprived areas. Several studies have noted gender, racial, and ethnic disparities in advanced medical therapies, including EVAR.25-27 Our findings add to the disparities literature by highlighting the association between a lower odds of EVAR and higher level of neighborhood deprivation, after controlling for race and gender. The mechanism of this association could be multifaceted. While we cannot exclude that bias toward the socioeconomically deprived plays a role in patient selection for endovascular therapies, it is also possible that environmental stresses associated with neighborhood deprivation may play a role in AAA formation and complexity, creating aortic anatomy not amenable to endovascular repair. Aneurysmal disease formation, growth, and rupture is a complex process with both genetic and environmental factors.28 For example, telomere length – a biomarker of cellular aging – has been shown to be significantly shortened in the endothelial cells, vascular smooth muscle cells, and blood lymphocytes from patients with AAA compared to controls.29 Further, a shorter telomere length, indicative of an older biological age, has been associated with lower socioeconomic status and financial instability.30 While we do not wish to draw a clear association between such complex biological processes as telomere length and aneurysm complexity in deprived areas, we do hope to display one of many possible biologic mechanisms and need for future studies of anatomic data and complexity. Of note, the significance of the association of EVAR with level of deprivation was not corrected for multiple hypothesis testing as we consider this outcome to be independent of presenting under age 65.
When examining the characteristics of young patients (<65 years), we found that those with prior knowledge of their AAA diagnosis that resided in the most deprived neighborhoods had a longer time from diagnosis to rupture (3.25 versus 0.8 years). This difference likely reflects the association between deprived neighborhoods and receipt of adequate medical care, subspecialty care, or comprehensive patient counseling regarding risks and benefits of aneurysm repair. For example, subspecialty medical care may be inaccessible to patients residing in limited resourced setting due to scheduling, transportation, or other well-documented barriers to care.31,32 This finding further supports the concept of enhanced educational or screening campaigns in under resourced neighborhoods to encourage repair for those who have knowledge of their own diagnosis as well as diagnostic opportunity for those who do not. However, the authors acknowledge that clinical severity on presentation may limit a patient’s ability to communicate any prior knowledge of their AAA and that this finding is observational to this study and further investigation is warranted.
Additional study limitations include the strong chance of residual confounding bias from unmeasured confounding variables along with those inherent to the ADI measure. Specifically, that the ADI does not have data on undocumented immigrants or those without a home address, and thus these people were not included in our study. Next, we assigned the ADI based upon the most recent version of the Neighborhood Atlas (2019), using the patients’ address location at the time of retrospective electronic medical record review, as the address at the time of repair was not available. Our data were also limited by lack of anatomic data or aneurysm complexity that may be related to the use of EVAR in deprived regions and information regarding family history of aneurysmal disease which relates to screening eligibility. Additionally, while smoking has been shown to associate with AAA disease in a dose-dependent fashion, our study lacks reliable knowledge of prior smoking exposure (i.e. pack-years, second-hand exposure) prior to rupture.33 Similarly, our study does not include the degree of control of comorbid conditions such as hypertension, nor does it include patients who presented with rAAA and did not undergo repair or patients who died outside of the hospital from a rAAA. And lastly, our findings were limited to Pennsylvania and surrounding states, a racially and ethnically homogenous region of the United States, and our results may need to be validated in other regions and nationally. Of note, there were no patients in our cohort who identified as Hispanic. Despite these, our large retrospective cohort study of patients who underwent repair for rAAA is novel, and attempts to understand how SDOH relate to one of the most devastating and preventable life-threatening emergencies in vascular surgery.
In this study of patients who underwent rAAA repair, residing in the most deprived neighborhoods was associated with a higher adjusted odds of presenting under age 65 years and receiving an open repair. Deprived neighborhoods should be populations of special interest for enhanced education, follow-up, and a younger AAA screening age. These findings stress the importance of developing strategies for early prevention and diagnosis of cardiovascular diseases among socioeconomically deprived patients given the earlier onset of devastating outcomes that stem from current healthcare inequities.
Supplementary Material
ARTICLE HIGHLIGHTS.
Type of Research:
Single center retrospective cohort study.
Key Findings:
Of 632 patients who underwent rAAA repair, residing in the most deprived neighborhoods was associated with presenting under the screening age of 65 years after adjusting for non-modifiable (OR 2.02;95% CI 1.39-2.95; p<0.001), and non-modifiable and modifiable (OR 2.22;95% CI 1.56-3.16; p<0.001) confounding variables compared to those from less-deprived neighborhoods.
Take home message:
The most deprived neighborhoods may be tangible populations of interest for a younger screening age or focused efforts to improve education and access to care.
ACKNOWLEDGMENT
Edith Tzeng MD, Chief of Vascular at the VA Pittsburgh Health System, and UPMC Professor of Surgery, was integral in the concept and design of this study and provided verbal input throughout the project.
FINANCIAL SUPPORT AND POTENTIAL CONFLICT OF INTEREST
This work was funded in part by the grant 5T32HL0098036 from the National Heart, Lung, and Blood Institute (Phillips, Reitz, Andraska), L30 AG064730 National Institute on Aging (Reitz), and the University of Pittsburgh holds a Physician-Scientist Institutional Award from the Burroughs Wellcome Fund (Andraska). These funding sources had no role in the design and conduct of the study; data collection, management, analysis, and/or interpretation; preparation, review, and/or approval of the article; or decision to submit for publication. The authors have no significant conflict of interest.
Footnotes
Presentation information: This work was presented as a full-length plenary talk during the Vascular Annual Meeting in San Diego, August 2021.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Department of Health and Human Services Centers for Medicare & Medicaid Services. Opportunities in Medicaid and CHIP to Address Social Determinants of Health (SDOH). 2021. p1–50. Available from: https://www.medicaid.gov/federal-policy-guidance/downloads/sho21001.pdf
- 2.Jolly S, Vittinghoff E, Chattopadhyay A, Bibbins-Domingo K. Higher cardiovascular disease prevalence and mortality among younger blacks compared to whites. Am J Med. 2010;123(9):811–8. [DOI] [PubMed] [Google Scholar]
- 3.Sidney S, Quesenberry CP, Jaffe MG, Sorel M, Nguyen-Huynh MN, Kushi LH, et al. Recent Trends in Cardiovascular Mortality in the United States and Public Health Goals. JAMA Cardiol. 2016;1(5):594–9. [DOI] [PubMed] [Google Scholar]
- 4.Soden PA, Zettervall SL, Deery SE, Hughes K, Stoner MC, Goodney PP, et al. Black patients present with more severe vascular disease and a greater burden of risk factors than white patients at time of major vascular intervention. J Vasc Surg. 2018;67(2):549–556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zommorodi S, Leander K, Roy J, Steuer J, Hultgren R. Understanding abdominal aortic aneurysm epidemiology: Socioeconomic position affects outcome. J Epidemiol Community Health. 2018;72(10):904–10. [DOI] [PubMed] [Google Scholar]
- 6.Vogel TR, Cantor JC, Dombrovskiy VY, Haser PB, Graham AM. AAA repair: sociodemographic disparities in management and outcomes. Vasc Endovasc Surg. 2008;42(6):555–60. [DOI] [PubMed] [Google Scholar]
- 7.Marcaccio CL, Schermerhorn ML. Epidemiology of abdominal aortic aneurysms. Seminars in Vascular Surgery. 2021;34(1):29–37. [DOI] [PubMed] [Google Scholar]
- 8.Chaikof EL, Dalman RL, Eskandari MK, Jackson BM, Lee WA, Mansour MA, et al. The Society for Vascular Surgery practice guidelines on the care of patients with an abdominal aortic aneurysm. J Vasc Surg. 2018;67(1):2–77. [DOI] [PubMed] [Google Scholar]
- 9.Owens DK, Davidson KW, Krist AH, Barry MJ, Cabana M, Caughey AB, et al. Screening for Abdominal Aortic Aneurysm: US Preventive Services Task Force Recommendation Statement. JAMA. 2019;322(22):2211–8. [DOI] [PubMed] [Google Scholar]
- 10.Dansey KD, Varkevisser RRB, Swerdlow NJ, Li C, de Guerre LEVM, Liang P, et al. Epidemiology of endovascular and open repair for abdominal aortic aneurysms in the United States from 2004 to 2015 and implications for screening. J Vasc Surg. 2021;74(2):414–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Berman AN, Biery DW, Ginder C, Singh A, Baek J, Wadhera RK, et al. Association of Socioeconomic Disadvantage with Long-term Mortality after Myocardial Infarction: The Mass General Brigham YOUNG-MI Registry. JAMA Cardiol. 2021;6(8):880–888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kind AJH, Jencks S, Brock J, Yu M, Bartels C, Ehlenbach W, et al. Neighborhood socioeconomic disadvantage and 30-day rehospitalization: A retrospective cohort study. Ann Intern Med. 2014;161(11):765–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kurani SS, McCoy RG, Lampman MA, Doubeni CA, Rutten LJF, Inselman JW, et al. Association of Neighborhood Measures of Social Determinants of Health With Breast, Cervical, and Colorectal Cancer Screening Rates in the US Midwest. JAMA Network Open. 2020;3(3):e200618–e200618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kind AJH, Buckingham WR. Making Neighborhood-Disadvantage Metrics Accessible — The Neighborhood Atlas. N Engl J Med. 2018;378(26):2456–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: Guidelines for reporting observational studies. International Journal of Surgery. 2014;12(12):1495–9.25046131 [Google Scholar]
- 16.Manson S, Schroeder J, Van Riper D, Kugler T, Ruggles S. IPUMS National Historical Geographic Information System: Version 16.0[dataset]. Minneapolis, MN: IPUMS. 2021. 10.18128/D050.V16.0 [DOI] [Google Scholar]
- 17.QGIS Development Team 2021. QGIS Geographic Information System. Open Source Geospatial Foundation Project. http://qgis.osgeo.org [Google Scholar]
- 18.Phillips AR, Reitz KM, Myers S, Thoma F, Andraska EA, Jano A, et al. Association Between Black Race, Clinical Severity, and Management of Acute Pulmonary Embolism: A Retrospective Cohort Study. J Am Heart Assoc. 2021;10:e021818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.VanderWeele TJ, Robinson WR. On the causal interpretation of race in regressions adjusting for confounding and mediating variables. Epidemiology. 2014;25(4):473–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schwamm LH, Reeves MJ, Pan W, Smith EE, Frankel MR, Olson D, et al. Race/Ethnicity, quality of care, and outcomes in ischemic stroke. Circulation. 2010;121(13):1492–501. [DOI] [PubMed] [Google Scholar]
- 21.Benson RA, Meecham L, Fisher O, Loftus IM Ultrasound screening for abdominal aortic aneurysm: current practice, challenges and controversies. Br J Radiol. 2018;91(1090). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ali MU, Fitzpatrick-Lewis D, Kenny M, Miller J, Raina P, Sherifali D. A systematic review of short-term vs long-term effectiveness of one-time abdominal aortic aneurysm screening in men with ultrasound. J Vasc Surg. 2018;68(2):612–23. [DOI] [PubMed] [Google Scholar]
- 23.Thompson SG, Bown MJ, Glover MJ, Jones E, Masconi KL, Michaels JA, et al. Screening women aged 65 years or over for abdominal aortic aneurysm: A modelling study and health economic evaluation. Health Technol Assess. 2018;22(43):V–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Altobelli E, Rapacchietta L, Profeta VF, Fagnano R. Risk Factors for Abdominal Aortic Aneurysm in Population-Based Studies: A Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2018;15(12). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yang Y, Lehman EB, ziz F. African Americans are less likely to have elective endovascular repair of abdominal aortic aneurysms. J Vasc Surg. 2019;70(2):462–70. [DOI] [PubMed] [Google Scholar]
- 26.Cruz-Flores S, Rabinstein A, Biller J, Elkind MSV, Griffith P, Gorelick PB, et al. Racial-ethnic disparities in stroke care: The American experience: A statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2011. Jul;42(7):2091–116. [DOI] [PubMed] [Google Scholar]
- 27.Majhail NS, Nayyar S, Santibañez ME, Murphy EA, Denzen EM. Racial disparities in hematopoietic cell transplantation in the United States. Bone Marrow Transplantation. 2012;47(11):1385–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kuivaniemi H, Platsoucas CD, Tilson MD 3rd. Aortic Aneurysms: an Immune Disease with a Strong Genetic Component. Circulation. 2008;117(2):242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cafueri G, Parodi F, Pistorio A, Bertolotto M, Ventura F, Gambini C, et al. Endothelial and Smooth Muscle Cells from Abdominal Aortic Aneurysm Have Increased Oxidative Stress and Telomere Attrition. PLoS ONE. 2012;7(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Needham BL, Salerno S, Roberts E, Boss J, Allgood KL, Mukherjee B. Do black/white differences in telomere length depend on socioeconomic status? Biodemography Soc Biol. 2019;65(4):287–312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Johnston KJ, Wen H, Maddox KEJ. Lack Of Access To Specialists Associated With Mortality And Preventable Hospitalizations Of Rural Medicare Beneficiaries. Health Affairs. 2019;38(12):1993–2002. [DOI] [PubMed] [Google Scholar]
- 32.Johnston KJ, Hammond G, Meyers DJ, Joynt Maddox KE. Association of Race and Ethnicity and Medicare Program Type With Ambulatory Care Access and Quality Measures. JAMA. 2021;326(7):628–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wilmink TBM, Quick CRG, Day NE. The association between cigarette smoking and abdominal aortic aneurysms. J Vasc Surg. 1999;30(6):1099–105. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.